Recent advances in deep learning have made it possible to visually track objects from a video sequence. Moreover, as transformers got introduced in computer vision, new state-of-the-art performances were achieved in visual tracking. However, most of these studies have used attentions to correlate the distinguishing factors between target-object and candidate-objects to localise the object throughout the video sequence. This approach is not adequate for tracking tiny objects. Also, conventional trackers in general are often not applicable to tracking extreme small objects, or objects that are moving fast. Therefore, the purpose of this study is to improve current methods to track tiny fast-moving objects, with the help of attentions. A deep neural network, named AATrackT, is built to address this gap by referring to it as a visual image segmentation problem. The proposed method is using data extracted from broadcasting videos of the sport Tennis. Moreover, to capture the global context of images, attention augmented convolutions are used as a substitute to the conventional convolution operation. Contrary to what the authors assumed, the experiment showed an indication that using attention augmented convolutions did not contribute to increasing the tracking performance. Our findings showed that the reason is mainly that the spatial resolution of the activation maps of 72x128 is too large for the attention weights to converge.

DEtection TRansformer, DETR, introduces an innovative design for object detection based on softmax attention. However, the softmax operation produces dense attention patterns, i.e., all entries in the attention matrix receive a non-zero weight, regardless of their relevance for detection. In this work, we explore several alternatives to softmax to incorporate sparsity into the architecture of DETR. Specifically, we replace softmax with a sparse transformation from the α-entmax family: sparsemax and entmax-1.5, which induce a set amount of sparsity, and α-entmax, which treats sparsity as a learnable parameter of each attention head. In addition to evaluating the effect on detection performance, we examine the resulting attention maps from the perspective of explainability. To this end, we introduce three evaluation metrics to quantify the sparsity, complementing the qualitative observations. Although our experimental results on the COCO detection dataset do not show an increase in detection performance, we find that learnable sparsity provides more flexibility to the model and produces more explicative attention maps. To the best of our knowledge, we are the first to introduce learnable sparsity into the architecture of transformer-based object detectors. / DEtection Transformer, DETR, introducerar en innovativ design för objektdetektering baserad på softmax attention. Softmax producerar tät attention, alla element i attention-matrisen får en vikt skild från noll, oberoende av deras relevans för objektdetektering. Vi utforskar flera alternativ till softmax för att inkludera gleshet i DETRs arkitektur. Specifikt så ersätter vi softmax med en gles transformation från α-entmax familjen: sparsemax och entmax1.5, vilka inducerar en fördefinierad mängd gleshet, och α-entmax, som ser gleshet som en träningsbar parameter av varje attention-huvud. Förutom att evaluera effekten på detekteringsprestandan, så utforskar vi de resulterande attention-matriserna från ett förklarbarhetsperspektiv. Med det som mål så introducerar vi tre olika metriker för att evaluera gleshet, som ett komplement till de kvalitativa observationerna. Trots att våra experimentella resultat på COCO, ett utmanande dataset för objektdetektering, inte visar en ökning i detekteringsprestanda, så finner vi att träningsbar gleshet ökar modellens flexibilitet, och producerar mer förklarbara attentionmatriser. Såvitt vi vet så är vi de första som introducerar träningsbar gleshet i transformer-baserade arkitekturer för objektdetektering.

En janvier 2008, la Commission européenne adopte le « Paquet Climat-Energie contraint les 27 États membres à atteindre trois objectifs d'ici à 2020 : réduction de 20% des émissions de gaz à effets de serre, augmentation de 20% de la part des énergies renouvelables dans le bouquet énergétique, et réduction de 20% de la consommation énergétique. Cettedirective se traduit dans les Etats membres par des programmes nationaux de développement des filières renouvelables. Dans ce cadre, notre réflexion porte sur les conditions de leur diffusion en France et au Royaume-Uni. Sous leurs formes décentralisées, celles-ci se heurtent à des freins techniques, matériels et symboliques qui leur confèrent un statut d’énergies « alternatives ». Cette recherche s’attache à comprendre les systèmes d’actions associés à la décision d’investissement, d’acquisition et d’intégration de quatre technologies renouvelables décentralisées dans la sphère domestique : les chauffe-eau solaires, les systèmes photovoltaïques, les pompes à chaleur, ainsi que les mini-éoliennes au Royaume-Uni. En procédant à une étude des filières, nous cherchons à identifier les conditions sociales entourant la diffusion de ces innovations technologiques auprès de ménages particuliers. Dans une perspective méso-sociale et micro-sociale, nous nous concentrons sur cinq groupes d’acteurs : individus appartenant aux sphères de l’Etat, du marché, de la société civile, ménages adoptants, et ménages non-adoptants ayant interrompu leur processus d’acquisition / In January 2008, the European Commission votes the "Climate and Energy Package", which forces the 27 Member States to achieve three goals by 2020: reduction in EU greenhouse gas emissions of at least 20% below 1990 levels; 20% of EU energy consumption to come from renewable resources, 20% reduction in primary energy use compared with projected levels, to be achieved by improving energy efficiency. This directive is reflected in the member states through national programs of development of renewable energy sources. In this context, our research focuses on the conditions of their diffusion in France and in the UK. As decentralized renewables, they have to cope with technical, material and symbolic barriers which give them an "alternative" status. This research seeks to analyze the systems of actions associated with the decision, the acquisition and the integration of four decentralized renewable technologies in the domestic space: solar thermal heating systems, photovoltaics, heat pumps, as well as mini-wind turbines in the UK. In analyzing these renewable networks, we seek to identify the social conditions surrounding the spread of these technological innovations to individual households. In a meso-social and a micro-social perspective, our research focuses on five groups of actors, belonging to the spheres of government, market, civil society, adopters and non-adopters who interrupted their process of acquisition

Este trabalho apresenta um modelo de acompanhamento e atendimento de pacientes crônicos baseado nos conceitos de relacionamento de clientes utilizados nas empresas, mais especificamente aqueles presentes na tecnologia CRM (Customer Relationship Management). A esse modelo denominou-se Gestão do Relacionamento com o Paciente Crônico (GRPC). A tecnologia CRM é uma estratégia com ferramentas para implementar um programa de relacionamento e fidelidade entre o cliente/consumidor e o fornecedor. O modelo GRPC, ao utilizar o conceito de CRM no atendimento ao paciente, apresenta uma estratégia de acompanhamento e monitoramento de pacientes crônicos diferente da abordagem tradicional, muitas vezes baseada somente no tratamento da doença. Da mesma forma que o CRM consegue atingir os clientes dos mais variados níveis através dos seus canais de comunicação, neste modelo propõe-se utilizar a mesma tecnologia a fim de garantir um acompanhamento efetivo e adequado a todas as camadas populacionais. Para implementar o modelo GRPC propõe-se a criação de centrais de relacionamento de pacientes crônicos, que compõem a infra-estrutura do modelo ao combinar, de maneira adequada, a troca de informações, as campanhas, a transmissão e o processamento de dados, com a finalidade de melhorar o relacionamento com o paciente, através da tecnologia de telefonia e computação. Para uma avaliação preliminar do modelo, elaborou-se uma central de monitoração de pacientes diabéticos e desenvolveu-se um sistema de monitoração para diabéticos. A modelagem da central de monitoração foi feita utilizando-se a extensão da UML para a modelagem de negócio, o que permitiu analisar a abrangência do modelo. O sistema de monitoração, denominado TeleDM, foi desenvolvido visando a realizar as simulações necessárias para os estudos desse trabalho. A partir da modelagem de negócio elaboraram-se os modelos para o desenvolvimento do sistema. Após a avaliação de suas características optou-se por implementá-lo utilizando a tecnologia de orientação a agentes e objetos, o que permitiu verificar que tal combinação é adequada na solução de sistemas complexos com as características do sistema TeleDM. / This paper shows a chronic patient follow-up and attendance model based on the concepts of customers’ relationship used in the companies, mainly those existent in CRM technology (Customer Relationship Management). The model was designated as Chronic Patient Relationship Management (CPRM). The CRM technology is a strategy with tools to implement a relationship and fidelity program between the client / customer and the supplier. The CPRM model, by using the CRM concept on patient attendance, presents a strategy to follow-up and monitoring the chronic patient different from the usual traditional approach, which many times only consists in illness treatment. In the same way as the CRM, this model is able to reach clients of every condition through its communication channels; and suggests the use of the same technology in order to guarantee an effective and suitable follow-up for all social layers. To implement the CPRM model, the creation of relationship centers for chronic patients, thus building the model’s infrastructure when properly connecting the information exchange, campaigns, and data processing and transmitting, for the purpose of improving the relationship with the patient through telephone and computing technology. For a preliminary assessment of the model, a monitoring center was elaborated for diabetic patients, as well as a system for diabetes monitoring. The modeling of the center was done using the UML extension for business modeling, which allowed analyzing the model’s coverage. The monitoring system known, as TeleDM was developed to execute the simulations needed for this work’s studies. The models for the system’s development were elaborated from the business modeling. After evaluating its features, the choice was implementing it with agent and object-oriented technology, thus allowing checking this combination suitability for complex systems solution with the TeleDM system features.

Este trabalho apresenta um modelo de acompanhamento e atendimento de pacientes crônicos baseado nos conceitos de relacionamento de clientes utilizados nas empresas, mais especificamente aqueles presentes na tecnologia CRM (Customer Relationship Management). A esse modelo denominou-se Gestão do Relacionamento com o Paciente Crônico (GRPC). A tecnologia CRM é uma estratégia com ferramentas para implementar um programa de relacionamento e fidelidade entre o cliente/consumidor e o fornecedor. O modelo GRPC, ao utilizar o conceito de CRM no atendimento ao paciente, apresenta uma estratégia de acompanhamento e monitoramento de pacientes crônicos diferente da abordagem tradicional, muitas vezes baseada somente no tratamento da doença. Da mesma forma que o CRM consegue atingir os clientes dos mais variados níveis através dos seus canais de comunicação, neste modelo propõe-se utilizar a mesma tecnologia a fim de garantir um acompanhamento efetivo e adequado a todas as camadas populacionais. Para implementar o modelo GRPC propõe-se a criação de centrais de relacionamento de pacientes crônicos, que compõem a infra-estrutura do modelo ao combinar, de maneira adequada, a troca de informações, as campanhas, a transmissão e o processamento de dados, com a finalidade de melhorar o relacionamento com o paciente, através da tecnologia de telefonia e computação. Para uma avaliação preliminar do modelo, elaborou-se uma central de monitoração de pacientes diabéticos e desenvolveu-se um sistema de monitoração para diabéticos. A modelagem da central de monitoração foi feita utilizando-se a extensão da UML para a modelagem de negócio, o que permitiu analisar a abrangência do modelo. O sistema de monitoração, denominado TeleDM, foi desenvolvido visando a realizar as simulações necessárias para os estudos desse trabalho. A partir da modelagem de negócio elaboraram-se os modelos para o desenvolvimento do sistema. Após a avaliação de suas características optou-se por implementá-lo utilizando a tecnologia de orientação a agentes e objetos, o que permitiu verificar que tal combinação é adequada na solução de sistemas complexos com as características do sistema TeleDM. / This paper shows a chronic patient follow-up and attendance model based on the concepts of customers’ relationship used in the companies, mainly those existent in CRM technology (Customer Relationship Management). The model was designated as Chronic Patient Relationship Management (CPRM). The CRM technology is a strategy with tools to implement a relationship and fidelity program between the client / customer and the supplier. The CPRM model, by using the CRM concept on patient attendance, presents a strategy to follow-up and monitoring the chronic patient different from the usual traditional approach, which many times only consists in illness treatment. In the same way as the CRM, this model is able to reach clients of every condition through its communication channels; and suggests the use of the same technology in order to guarantee an effective and suitable follow-up for all social layers. To implement the CPRM model, the creation of relationship centers for chronic patients, thus building the model’s infrastructure when properly connecting the information exchange, campaigns, and data processing and transmitting, for the purpose of improving the relationship with the patient through telephone and computing technology. For a preliminary assessment of the model, a monitoring center was elaborated for diabetic patients, as well as a system for diabetes monitoring. The modeling of the center was done using the UML extension for business modeling, which allowed analyzing the model’s coverage. The monitoring system known, as TeleDM was developed to execute the simulations needed for this work’s studies. The models for the system’s development were elaborated from the business modeling. After evaluating its features, the choice was implementing it with agent and object-oriented technology, thus allowing checking this combination suitability for complex systems solution with the TeleDM system features.